#333

Excuse me if I insist on this article. I’ve asked the Deepseek AI for a critical review of it, potential biases and inchoerencies.

Based on a critical review of the attached article (Optimal dietary patterns for healthy aging, Tessier et al., 2025) and comparison with existing literature, the following biases and inconsistencies are discernible:

1. Potential Biases

a. Selection Bias

  • Population homogeneity: The study uses data from the Nurses’ Health Study (NHS) and Health Professionals Follow-Up Study (HPFS), which primarily include health professionals (mostly women in NHS and men in HPFS) of predominantly European ancestry (95% in pooled cohorts). This limits generalizability to diverse populations (e.g., non-European ancestry, lower socioeconomic status).
  • Healthy cohort effect: Participants were free of chronic diseases at baseline, potentially excluding individuals with diet-related comorbidities, which may skew results toward healthier aging outcomes.

b. Measurement Bias

  • Self-reported dietary data: Dietary patterns were assessed via food frequency questionnaires (FFQs), which are prone to recall bias and misreporting (e.g., underreporting unhealthy foods).
  • Healthy aging domains: Cognitive, physical, and mental health assessments relied on self-reported questionnaires (e.g., Subjective Cognitive Decline, Geriatric Depression Scale), which are less objective than clinical evaluations.

c. Confounding Bias

  • Residual confounding: While the study adjusted for lifestyle factors (e.g., smoking, BMI, physical activity), unmeasured confounders (e.g., genetic predisposition, environmental exposures, stress) could influence associations.
  • Socioeconomic status (SES): SES was adjusted via a composite score, but residual socioeconomic disparities (e.g., access to healthcare, education) may persist.

d. Funding and Conflict of Interest

  • Industry influence: One author (M.G.-.) received funding from the International Nut Council, which may introduce bias toward diets emphasizing nuts (a component of several studied patterns).

2. Inconsistencies with Existing Literature

a. Weaker Association of Healthful Plant-Based Diet (hPDI)

  • The study found weaker associations between the healthful plant-based diet (hPDI) and healthy aging compared to other patterns (e.g., AHEI, Mediterranean). This contrasts with prior studies emphasizing strict plant-based diets for longevity (e.g., Satija et al., 2017).
  • Explanation: The hPDI negatively weights all animal-based foods, potentially excluding beneficial components (e.g., fish, low-fat dairy). This aligns with critiques that overly restrictive plant-based diets may miss synergistic nutrients (e.g., vitamin B12, omega-3s).

b. Sex-Specific Differences

  • Stronger associations in women (e.g., AHEI OR = 1.74 vs. 1.58 in men) conflict with some studies reporting similar benefits across sexes (e.g., Zhou et al., 2021).
  • Possible cause: Hormonal or metabolic differences (e.g., menopause effects) may modulate dietary impacts, but the study does not explore mechanisms.

c. Ultraprocessed Foods (UPFs)

  • The inverse association between UPFs and healthy aging aligns with global consensus (e.g., Monteiro et al., 2019). However, the NOVA classification used here has been criticized for oversimplifying food processing levels (e.g., grouping minimally processed nuts with industrially altered foods).

d. Planetary Health Diet Index (PHDI)

  • The PHDI showed strong associations with cognitive health but weaker links to chronic disease prevention. This contrasts with its original design (Willett et al., 2019), which emphasizes both health and sustainability.

3. Methodological Limitations

  • Survivorship bias: Only participants surviving to 70+ years were included, excluding those who died earlier (potentially with worse diets).
  • Temporality: Diet was assessed midlife (1986–2010), but dietary habits may change with aging, limiting causal inference.
  • Multiple comparisons: Subgroup analyses (e.g., sex, BMI) increase Type I error risk without Bonferroni correction.

4. Alignment with Consensus

The study aligns broadly with evidence that diets rich in plant foods, unsaturated fats, and fish (e.g., Mediterranean, DASH) promote healthy aging. However, its emphasis on moderate animal-food inclusion (e.g., low-fat dairy in AHEI) diverges from strict plant-based recommendations.

Recommendations for Interpretation

  1. Generalizability: Apply findings cautiously to non-European, lower-SES populations.
  2. Dietary nuance: Prioritize patterns like AHEI or Mediterranean over restrictive plant-based diets (hPDI).
  3. Mechanistic research: Investigate sex-specific pathways and the role of food processing.

This study advances dietary guidelines but requires replication in diverse cohorts and with objective health measures. :seedling::microscope:.

#334

I think that’s a good idea. I don’t know why he’d be biased against salt though, for example.

#335

More ruminations on this study. what if we want to avoid altogether personal bias and rely on the logical inference of AI? I asked to build a daily omnivorous menu according to the results of the article, what comes out seems overly interesting.

Based on the findings of the study (Optimal dietary patterns for healthy aging, Tessier et al., 2025), here’s a daily omnivore menu designed to optimize healthy longevity, emphasizing foods linked to intact cognitive, physical, and mental health, survival to 70+ years, and reduced chronic disease risk:

Breakfast

  • Oatmeal bowl:
    • Steel-cut oats cooked with unsweetened almond milk.
    • Topped with blueberries, walnuts, and chia seeds (rich in fiber, antioxidants, omega-3s).
    • Side of low-fat Greek yogurt (probiotics, calcium).
    • Optional: Drizzle of honey (≤1 tsp).

Mid-Morning Snack

  • Apple slices with almond butter (fiber, healthy fats).
  • Green tea (polyphenols for cognitive health).

Lunch

  • Mediterranean-inspired salad:
    • Base: Mixed greens (spinach, kale), cherry tomatoes, cucumbers, roasted chickpeas, olives.
    • Protein: Grilled salmon (omega-3 fatty acids) or skinless chicken breast (lean protein).
    • Whole grain: Quinoa or farro.
    • Dressing: Extra-virgin olive oil, lemon juice, turmeric, black pepper.

Afternoon Snack

  • Crudités: Carrot sticks, bell peppers, and broccoli with hummus (fiber, plant protein).
  • Handful of mixed nuts (almonds, pistachios).

Dinner

  • Grilled sardines or mackerel (omega-3s, vitamin D) or tofu stir-fry (plant-based option).
  • Steamed vegetables: Broccoli, Brussels sprouts, asparagus (sulforaphane for detoxification).
  • Whole grain: Brown rice or barley.
  • Side: Lentil soup (plant protein, iron).

Dessert

  • Dark chocolate (≥70% cocoa, 1-2 squares) + orange slices (vitamin C, flavonoids).

Key Principles

  1. Plant-forward: ≥50% of calories from fruits, vegetables, whole grains, nuts, legumes.
  2. Healthy fats: Olive oil, avocado, nuts, fatty fish (anti-inflammatory).
  3. Lean proteins: Fish (2–3x/week), poultry, legumes; limit red meat to ≤1x/week.
  4. Low-fat dairy: Yogurt or kefir for gut health.
  5. Minimize:
    • Processed meats, sugary drinks, refined grains.
    • Sodium (<2,300 mg/day; use herbs/spices instead).

Rationale

  • Vegetables & fruits: Associated with 1.92x odds of healthy aging (Fig. 4).
  • Whole grains: Linked to reduced chronic disease risk (OR 1.45–1.86).
  • Fish/nuts: Omega-3s support cognitive health and lower inflammation (EDIP/rEDIH scores).
  • Legumes: Improve insulin sensitivity (EDIH reversal).

Hydration

  • Water, herbal teas, black coffee (antioxidants).
  • Optional: 1 glass red wine (5 oz) with dinner (AHEI component; polyphenols).

This menu aligns with the AHEI, Mediterranean, and DASH patterns, which showed the strongest associations with healthy aging in the study. Adjust portions to meet individual caloric needs (e.g., ~2,000 kcal/day for average adults). :green_salad::fish::ear_of_rice:

#336

He evidently assigns a high degree of reliability to the DASH protocol, which si also described in the article with ORs in Figure 2 and appears to be very much correlated to the results of the MNEditerranean protocol

image
image1332×941 282 KB

#337

Greger is more radical, he suggests no added sodium consumption, based on an argument to prevent increase in risk of HBP with age based on ecological studies.

#338

I do not criticize Dr. Greger for such suggestions, which are generally healthy; rather, he often omits to state the evidence contrary to a pure vegan regimen, and people notice that, he loses credibility.

Back to sodium, he also suggests the alternative of KCl, which is very useful to limit sodium intake and increase potassium. To eliminate added sodium completely is not easy; it may also limit the palatability of vegetables and, as a consequence, limit the consumption of such a very healthy food group.

1 Like